HYDROGEN DIRECT INJECTION IN RECIPROCATING ENGINES USING COMMERCIAL INJECTORS

(1)

UNIVERSITÀ DI PISA

Scuola di Dottorato “Leonardo da Vinci”

PhD programme in Land Vehicles and Transport Systems

PhD Dissertation

HYDROGEN DIRECT INJECTION IN

RECIPROCATING ENGINES USING

COMMERCIAL INJECTORS

Alessio Simi

SSD ING-IND/08

2011

(2)

(3)

UNIVERSITÀ DI PISA

Scuola di Dottorato “Leonardo da Vinci”

PhD programme in Land Vehicles and Transport Systems

PhD Dissertation

HYDROGEN DIRECT INJECTION IN

RECIPROCATING ENGINES USING

COMMERCIAL INJECTORS

PhD Student: Alessio Simi

PhD Tutor: Prof. Luigi Martorano

SSD ING-IND/08

2011

(4)

(5)

(6)

(7)

Abstract

Several techniques are being developed to solve the problem of urban pollution. Among other solutions (improvements in engine control and combustion, electrical propulsion) one of the possibility is the employment of low carbon content fuels.

A nearly zero emission vehicle may be propelled by hydrogen, in this case the only polluting agents are nitrogen oxides when using internal combustion engines.

Though fuel cells are considered to be the most promising solution in the long term, they are still in the prototypical phase, while the internal combustion engine continues to be a relevant topic.

Hydrogen used as a fuel, combined with hybridization was considered to be a very eﬀective solution for urban vehicles such as small buses or light duty vehicles.

In particular, series hybrid vehicles can use very small displacement engines as range extenders resulting in a noticeable fuel economy when considering urban cycles.

This document is the Doctor of Philosophy course thesis in Land Vehicles and Transport System.

The object of the thesis is the study of a hydrogen direct injection system for recipro-cating engines using commercial injectors.

In particular, this study is applied to a 505cm3 single-cylinder engine. This is a typical unit displacement and consequently it will be possible to extend the results to multi-cylinder engines for automotive purpose.

This activity is a part of a project funded by the Tuscan region and carried out in cooperation with several Departments of the University of Pisa.

(8)

(9)

Contents

List of Figures

2.1 Car designed by Rivaz - 1807 . . . 6

2.2 Car designed by Etienne Lenoir - 1860 . . . 6

2.3 Vehicle designed by Norsk Hydro - 1933 . . . 7

2.4 Musashi 3 car - 1977 . . . 7

2.5 HICE Chevrolet Silverado/GMC Sierra 1500HD . . . 9

2.6 Hydrogen Toyota Prius . . . 9

2.7 BMW Hydrogen 7 . . . 10

2.8 Hydrogen Toyota Prius . . . 11

2.9 Ford P2000 . . . 12

2.10 Ford E-450 shuttle bus . . . 12

3.1 Chemical structure of common fuels . . . 17

3.2 Variation of hydrogen ﬂammability limits with temperature . . . 20

3.3 Flammability ranges of comparative fuels at atmospheric temperature . . . 21

4.1 Lombardini 15LD500 engine . . . 28

4.2 Lombardini 15LD500 engine - Curves . . . 30

4.3 Lombardini 15LD500 engine - Technical drawings . . . 31

4.4 Preliminary study - Air and hydrogen mass ﬂow rate . . . 32

5.1 1D diesel engine - One-dimensional model . . . 34

5.2 1D diesel engine - Comparison of brake torque . . . 35

5.3 1D diesel engine - Comparison of brake power . . . 35

5.4 1D diesel engine - Comparison of exhaust gas temperature . . . 36

5.5 1D diesel engine - Comparison of BSFC . . . 36

5.6 1D diesel engine - Air mass ﬂow rate . . . 37

5.7 1D hydrogen engine - Features AMESim injector . . . 38

5.8 1D hydrogen engine - Injection timing submodel . . . 39

5.9 1D hydrogen engine - Non-dimensional hydrogen mass ﬂow rate . . . 41

5.10 1D hydrogen engine - Critical mass ﬂow submodel . . . 41

5.11 1D hydrogen engine - Variable mass ﬂow submodel . . . 41

5.12 1D hydrogen engine - Trapezoidal proﬁle for the injection . . . 42

5.13 1D hydrogen engine - AMESim sub model used . . . 42

5.14 1D hydrogen engine - Injection timing and intake valve lift . . . 43 xi

(12)

5.15 1D hydrogen engine - Hydrogen modiﬁed injector . . . 44

5.16 1D hydrogen engine - Eﬀective hydrogen mass ﬂow rate . . . 44

5.17 1D hydrogen engine - Air mass ﬂow rate . . . 45

5.18 1D hydrogen engine - Hydrogen mass ﬂow rate . . . 45

5.19 1D hydrogen engine - Air mass for each cycle . . . 46

5.20 1D hydrogen engine - Hydrogen mass for each cycle . . . 46

5.21 1D hydrogen engine - Hydrogen injected mass ﬂow rate . . . 47

5.22 1D hydrogen engine - Sensitivity analysis on the injection timing . . . 48

5.23 0D injection system - Bigas numerical model . . . 48

5.24 0D injection system - Bigas plunger lift comparison . . . 49

5.25 0D injection system - Air mass ﬂow rate comparison . . . 49

5.26 0D injection system - Synerject Strata injector numerical model . . . 50

5.27 0D injection system - Hydrogen mass ﬂow rate analysis . . . 50

5.28 Injection system - Test equipment . . . 51

5.29 Injection system - Experimental instrumentation scheme . . . 51

5.30 0D Injection system - Injection system numerical model . . . 52

5.31 Injection system - Experimental hydrogen mass ﬂow rate . . . 52

5.32 0D injection system - Hydrogen mass flow rate comparison with 0 ms delay 53 5.33 0D injection system - Hydrogen mass flow rate comparison with 2.5 ms delay 54 5.34 0D injection system - Hydrogen mass flow rate comparison with 5 ms delay 54 5.35 Injection system - Injection pulses with 10 ms width and 0 ms delay . . . . 55

5.36 Injection system - Injection pulses with 10 ms width and 2.5 ms delay . . . 55

5.37 Injection system - Injection pulses with 10 ms width and 5 ms delay . . . . 56

6.1 CAD model - Original head of the engine . . . 58

6.2 CAD model - Modiﬁed head of the engine . . . 58

6.3 CAD model - Modiﬁed head with pressure sensor and spark plug . . . 59

6.4 CAD model - Injection system scheme . . . 59

6.5 CAD model - Modiﬁed head with injection system . . . 60

6.6 CAD model - Modiﬁed head and injection system scheme . . . 60

6.7 CAD model - Combustion chamber volume . . . 61

6.8 Exhaust Gas Recirculation scheme . . . 62

6.9 Prototype - Components of the hydrogen engine . . . 64

6.10 Prototype - Modiﬁed head with pressure sensor and spark plug . . . 64

6.11 Prototype - Bigas valves restrain block . . . 65

6.12 Prototype - Bigas valves, hydrogen line interface and restrain block - View 1 65 6.13 Prototype - Bigas valves, hydrogen line interface and restrain block - View 2 66 6.14 Prototype - Synerject injectors restrain block . . . 66

6.15 Prototype - Synerject injectors and restrain block - View 1 . . . 67

6.16 Prototype - Synerject injectors and restrain block - View 2 . . . 67

6.17 Prototype - Modiﬁed head and injection system - View 1 . . . 68 xii

(13)

List of Figures

6.18 Prototype - Modiﬁed head and injection system - View 2 . . . 68

6.19 Prototype - Modiﬁed head and injection system - View 3 . . . 69

6.20 Prototype - Modiﬁed piston . . . 69

6.21 Prototype - Assembled engine . . . 70

6.22 Prototype - The intake manifold and the mixing point . . . 71

6.23 Prototype - The heat exchanger and the gate valve . . . 72

6.24 Prototype - The injection system and the muﬄer . . . 72

7.1 Prototype - Several sensors applied to the engine . . . 75

7.2 Prototype - Several sensors applied to the engine head . . . 76

7.3 Scheme of the sensors applied to the engine . . . 76

7.4 Example of P-V diagram for a four-stroke cycle engine . . . 79

7.5 Results - Spark timing variations,φ= 0.8 . . . 82

7.6 Results - Spark timing variations,φ= 0.6 . . . 84

7.7 Results - Spark timing variations,φ_{= 0.4 . . . 85}

7.8 Results - Spark timing variations - Indicated in-cylinder pressure versus crankshaft degree,φ= 0.8, 3000 rpm . . . 86

7.9 Results - Spark timing variations - Indicated P-V diagram,φ= 0.8, 3000 rpm 86 7.10 Results - Spark timing variations - Indicated in-cylinder pressure versus crankshaft degree,φ= 0.8, 2000 rpm . . . 87

7.15 Results - Spark timing variations - Indicated P-V diagram,φ= 0.4, 3000 rpm 89 7.16 Results - Injection timing variations,φ= 0.8 . . . 91

7.17 Results - Injection timing variations,φ= 0.6 . . . 92

7.18 Results - Injection timing variations,φ= 0.4 . . . 93

7.19 Results - Injection timing variations - Indicated in-cylinder pressure versus crankshaft degree,φ= 0.8, 3000 rpm . . . 94

7.20 Results - Injection timing variations - Indicated P-V diagram,φ= 0.8, 3000 rpm . . . 94

(14)

7.24 Results - Injection timing variations - Indicated P-V diagram,φ= 0.6, 3000

rpm . . . 96

7.26 Results - Injection timing variations - Indicated P-V diagram,φ_{= 0.6, 2000} rpm . . . 97

7.29 Results -N Oxemissions map . . . 100

7.30 Results - Eﬀects of the hot EGR . . . 102

7.31 Results - Eﬀects of the cool EGR . . . 103

7.32 Results - Performance map . . . 104

7.33 Results - Eﬃciency map . . . 104

7.34 Results - Comparison between diesel and hydrogen engine . . . 106

A.1 Layout - Modiﬁed head . . . 113

A.2 Layout - Synerject injectors restrain block . . . 115

A.3 Layout - Bigas valves restrain block . . . 117

A.4 Layout - Bigas valves assembly . . . 119

A.5 Layout - Hydrogen line interface . . . 121

A.6 Layout - Pressure sensor interface . . . 123

A.7 Layout - Restrain block threaded fastener . . . 125

A.8 Layout - Assembly of the head of the engine . . . 127

A.9 Layout - Modiﬁed piston bowl . . . 129

(15)

List of Tables

2.1 Hydrogen vehicles overview . . . 13

3.1 Hydrogen properties compared with methane and iso-octane properties . . 18

3.2 Vapor and liquid densities of comparative substances . . . 18

3.3 Mixture properties of hydrogen–air, methane–air and iso-octane–air . . . . 18

3.4 Higher and lower heating values of comparative fuels . . . 23

3.5 Energy densities of comparative fuels . . . 24

3.6 Flashpoint of comparative fuels . . . 24

3.7 Auto-ignition temperature of comparative fuels . . . 25

3.8 Octane number of comparative fuels . . . 25

4.1 Lombardini 15LD500 speciﬁcations . . . 29

7.1 Eﬀects of EGR onη, BMEP, BSFC and N Oxemissions . . . 101

(16)

(17)

Chapter 1. Introduction

Chapter 1 Introduction

The current way of providing the world’s energy demand, based primarily on fossil fuel, is becoming increasingly untenable. Fossil fuel reserves, once hardly ever given a second thought, are now clearly exhaustible. Fossil fuel prices have never been more volatile, inﬂuenced ﬁrstly by economic acceleration mostly in China and India and subsequently by economic recession.

The diﬃculty of controlling prices and the uncertain reserves are strong incentives for pursuing energy security. Global warming and local pollution hot spots associated with fossil fuel usage are further signiﬁcant environmental and societal problems.

These are strong drivers for research, development and demonstrations of alternative energy sources, energy carriers, and in the case of transportation, powertrains. The use of hydrogen as an energy carrier is one of the options put forward in most governmental strategic plans for a sustainable energy system. The United States Department of Energy, the European Commission’s Directorate-General for Research, the Japanese Ministry of Economy, Trade and Industry, the Indian Ministry of New and Renewable Energy and many others have formulated vision reports and published funding calls for hydrogen programs [1]-[4].

The attractiveness of hydrogen lies in the variety of methods to produce hydrogen as well as the long-term viability of some of them (from fossil fuels, from renewable energy: biomass, wind, solar [5], from nuclear power etc.), the variety of methods to produce energy from hydrogen (internal combustion engines, gas turbines, fuel cells), virtually zero harmful emissions and potentially high eﬃciency at the point of its use.

Compared to biofuels, a recent study reported the yield of ﬁnal fuel per hectare of land for diﬀerent biomass derived fuels, and of hydrogen from photovoltaics or wind power [6]. The results show that the energy yield of land area is much higher when it is used to capture wind or solar energy. Compared to electricity, using hydrogen as an energy carrier is advantageous in terms of volumetric and gravimetric energy storage density. However, there are also serious challenges to overcome when hydrogen is to be used as an energy carrier.

Although better than batteries in storage terms, its very low density implies low energy densities compared to the fuels in use today, even when compressed to 700 bar or liqueﬁed,

(18)

both of which incur substantial energy losses. Thus distribution, bulk storage and onboard vehicle storage are heavily compromised. Also, in case of hydrogen-fueled vehicles, care must be taken to ensure that the well-to-wheel greenhouse gas emission reduction compared to hydrocarbon fuel turns out to be positive [7]. Nevertheless, the advantages oﬀered by hydrogen are signiﬁcant enough to warrant the exploration of its possibilities.

1.1 Hydrogen as fuel in internal combustion engines

There are numerous works and opinions as to what constitutes the ‘best’ fuel/ energy carrier as well as powertrain: heavily optimized hydrocarbon fueled engines, biofuels, electricity, hydrogen, etc. However, there are always a multitude of aspects to be taken into account, ranging from well-to-wheel (or cradle-to-grave) primary energy use, greenhouse gas emissions, tailpipe emissions relating to local pollution, cost, practicality, to customer acceptance, etc., which clearly are not all easily scored and ranked, making it very hard to predict the winner. Moreover, there appears to be no silver bullet, and numerous choices merit a detailed study in exploring the possibilities and drawbacks.

Concerning hydrogen as an energy carrier, Shelef and Kukkonen [7] have compared hydrogen fuel cell (H₂F C) vehicles and H₂ICE vehicles to gasoline vehicles and electric

vehicles, on the basis of well-to-wheel carbon dioxide (CO₂) emissions and primary energy use. Fifteen years have passed since that study, but overall the arguments and challenges for the implementation of hydrogen as a vehicle fuel largely hold. However, the data that were used clearly need updating to see how this aﬀects the conclusions. Shelef and Kukkonen concluded that theH₂F C vehicle could decrease primary energy use and

greenhouse gas emissions compared to gasoline and natural gas vehicles, but thatH₂ICE

vehicles increased both.

However, a recent study at Argonne National Laboratory that compares the fuel economy potential of hydrogen powertrains to conventional gasoline vehicles concludes that by 2045 aH₂ICE hybrid-electric vehicle (HEV) would only consume 9% more than

a H₂F C HEV, as a result of the recent and expected future signiﬁcant improvements

in hydrogen engine technology [8]. Thus, the perceived large diﬀerence in fuel economy between a H₂F C HEV and H₂ICE HEV is lower than frequently reported and will

decrease over time.

Currently,H₂ICEs are much cheaper than H₂F Cs, both directly and in terms of fuel

cost (with high fuel purity requirements for theH₂F Cs). Furthermore, using ICEs allows

bi-fuel operation (e.g., the engine can run on gasoline as well as on hydrogen), alleviating fuel station density and autonomy requirements. This could facilitate the start-up of a hydrogen infrastructure, where the experience gained with transport, fueling and storage directly translates to fuel cell vehicles.

This explains the U.S. Department of Energy’s position that while FC vehicles consis-tently achieve the highest fuel eﬃciency, theH₂ICE can serve as bridging technology and

(19)

3 Chapter 1. Introduction Similarly, the well-to-wheels study of the European Commission’s Directorate General Joint Research Center, in cooperation with CONCAWE and EUCAR, concludes that

(20)

(21)

Chapter 2. Hydrogen internal combustion engine vehicles

Chapter 2 Hydrogen internal combustion engine

vehicles

Hydrogen internal combustion engines for automotive application are intended to power vehicles and provide an equivalent level of drivability, range and safety as conventional-fuel vehicles. However, mainly due to the challenges of on-board hydrogen storage, current hydrogen-powered internal combustion engine vehicles have a limited range and in some cases reduced trunk space available compared to their conventional-fuel counterparts.

Nonetheless, due to the immediate availability of hydrogen combustion engines, the extensive knowledge in engine production, durability and maintenance as well as the capability of combustion engines to run on both hydrogen as well as conventional fuels (in most cases, gasoline), they are considered a bridging technology towards a widespread hydrogen infrastructure [1]. In this role, hydrogen internal combustion engine vehicles can be considered early adopters to help establishing and expanding a hydrogen infrastructure and building public awareness.

Numerous hydrogen engine-powered vehicles ranging from two-wheelers to passenger cars, pickup trucks to buses and oﬀ-road equipment have been designed, built and tested over the last decades. The following chapter is limited to selected hydrogen internal combustion engine vehicles; design studies and show-cars are excluded from this overview.

2.1 History

The concept of operating an internal combustion engine on hydrogen is almost as old as the internal combustion engine itself.

In 1807, François Isaac de Rivaz of Switzerland invented an internal combustion engine that used a mixture of hydrogen and oxygen for fuel. Rivaz designed a car for this engine (ﬁgure 2.1), the ﬁrst internal combustion-powered automobile [10].

Patented by Jean Joseph Etienne Lenoir in 1860, a gas-driven two-stroke engine with horizontal arrangement is considered the ﬁrst successful internal combustion engine (ﬁgure 2.2). The engine was powered by hydrogen generated via the electrolysis of water [11].

(22)

Figure 2.1: Car designed by Rivaz - 1807

(23)

7 Chapter 2. Hydrogen internal combustion engine vehicles As early as 1933, Norsk Hydro operated an internal combustion engine vehicle on hydrogen (ﬁgure 2.3) produced from on-board reforming of ammonia [12].

The first hydrogen DI engine dates back to 1933 when Erren Engineering Company proposed injecting slightly pressurized hydrogen into air or oxygen inside the combustion chamber rather than feeding the air–fuel mixture via a carburetor into the engine, a method that commonly resulted in violent backfiring. The patented system required special fuel injection and control mechanisms but left the other engine components intact. With hydrogen used as a booster, the system eliminated backfiring and achieved much better combustion of hydrocarbons with higher output and lower specific fuel consumption [13]. In 1974, Musashi Institute of Technology introduced the first Japanese hydrogen-fueled vehicle, called Musashi 1, using a 4-stroke hydrogen engine and high-pressure storage [14]. The Musashi 2, introduced in 1975, was equipped with hydrogen manifold injection on a 4-stroke engine in combination with liquid hydrogen storage [14]. In 1977, Musashi 3 (figure 2.4) was presented using a spark-ignited 2-stroke engine with hydrogen DI [15].

Figure 2.3: Vehicle designed by Norsk Hydro - 1933

Figure 2.4: Musashi 3 car - 1977

BMW in collaboration with DLR introduced their ﬁrst hydrogen vehicle in 1979.

2.2 Hydrogen vehicles characterization

Hydrogen internal combustion engine vehicles can be characterized as either conversion vehicles or dedicated vehicles, with conversion vehicles adapted for hydrogen operation

(24)

by either a vehicle manufacturer or an after-market supplier, whereas dedicated hydrogen cars are speciﬁcally designed and built for hydrogen operation by an original equipment manufacturer (OEM).

Hydrogen cars have also been built for mono-fuel operation with hydrogen as the only fuel as well as bi-fuel solutions with hydrogen as well as gasoline as fuel options. Based on the hydrogen on-board storage system, hydrogen cars can be grouped as compressed hydrogen and cryogenic liquid hydrogen vehicles. Hydrogen as an engine fuel has been applied to reciprocating internal combustion engines as well as rotary engines. The following sections give a brief overview of selected hydrogen vehicles. Automobiles that use hydrogen as a combustion enhancer in combination with another fuel are not considered in this overview.

2.3 Conversion vehicles

An example for a conversion truck with compressed hydrogen storage is the ET EC H₂ICE Truck Conversion based on a Chevrolet/GMC Truck Silverado/Sierra 1500HD

Crew Cab 2WD LS converted to hydrogen operation by Electric Transportation Engineering Corporation (figure 2.5). The 6-seated light-duty pickup truck is powered by a 6.0 LV-8 engine with hydrogen port-fuel injection. A belt-driven supercharger in combination with an intercooler is used to increase the power output of the engine. Hydrogen is stored in three 150 L, Type 3 (aluminum lined, carbon-fiber reinforced) tanks at a storage pressure of up to 350 bar, which results in approximately 10.5 kg of usable fuel. The vehicle has an estimated curb weight of 3000 kg [16]. A performance, emissions and fuel economy study of this vehicle at different air fuel ratios (2 < λ < 2.85; 0.50 < φ < 0.35) showed fuel consumption numbers between 4.1 and 4.5 kg of hydrogen per 100 km which is energy equivalent to 15.5 and 17 L of gasoline per 100 km atN Oxemissions levels in the Ultra

Low Emissions Vehicle (ULEV) and Super Ultra Low Emissions Vehicle (SULEV) ranges [17]. So far about 20ET EC H₂ICE Truck Conversion vehicles have been built.

Quantum Tecstar has converted over 30 vehicles to hydrogen operation using the Toyota Prius hybrid as a platform (ﬁgure 2.6). Two compressed hydrogen tanks replace the conventional gasoline tank, leaving the interior of the vehicle unchanged. The converted Prius engine is turbocharged in order to increase the power output in hydrogen operation. With a drivability similar to the gasoline counterpart, the Quantum Hydrogen Prius has an estimated range of 100–130 km per ﬁll while meeting SULEV emissions standards [18].

2.4 Bi-fuel vehicles

Since 1979, BMW has introduced six generations of hydrogen-powered internal combustion engine vehicles. The latest generation is the BMW Hydrogen 7 bi-fuel (ﬁgure 2.7), a luxury sedan powered by a 6.0 LV12 engine. According to the manufacturer’s claims, the BMW Hydrogen 7 vehicle has successfully completed the process of series development, meaning

(25)

9 Chapter 2. Hydrogen internal combustion engine vehicles

Figure 2.5: HICE Chevrolet Silverado/GMC Sierra 1500HD

(26)

that the vehicle and all components have gone through the same design, manufacturing and quality control processes as any other BMW vehicle. The new hydrogen model is built at BMW’s Dingolﬁng Plant (Germany) parallel to the other models in the BMW 7, 6 and 5 Series, with the drive unit in the BMW Hydrogen 7 coming like all BMW twelve-cylinder engines from the BMW engine production plant in Munich (Germany). The engine is equipped with two separate fuel systems allowing the vehicle to operate on gasoline as well as hydrogen. Gasoline is injected directly into the combustion chambers; hydrogen is injected into the intake manifolds of the naturally aspirated engine [19]. The vehicle is equipped with a cryogenic hydrogen tank located in the trunk of the vehicle in addition to the conventional gasoline tank. The cryogenic tank holds about 8 kg of liquid hydrogen which allows an estimated range of 200 km in hydrogen operation and another 480 km on gasoline [20]. Approximately 100 BMW Hydrogen 7 bi-fuel vehicles were built.

Since 1991, Mazda has developed several generations of hydrogen-powered rotary engine vehicles with the Mazda RX-8 Hydrogen RE being the most recent one unveiled in 2003 (ﬁgure 2.8). The hydrogen version of the Renesis engine is equipped with an electric-motor-assist turbocharger that is used to maximize the eﬀectiveness of forced induction throughout the engine speed range [21]. The most recent generation is equipped with two compressed hydrogen tanks with an operating pressure of up to 350 bar, giving the vehicle a range of approximately 100 km in hydrogen operation plus and additional 550 km on gasoline. A combination of lean and stoichiometric hydrogen combustion operation results in a 23% improvement in fuel economy compared to gasoline operation. The performance of the vehicle meeting Japanese SULEV standards is reduced from 154 kW in gasoline to 80 kW in hydrogen operation [22].

Figure 2.7: BMW Hydrogen 7

2.5 Dedicated hydrogen vehicles

The BMW Hydrogen 7 Mono-Fuel demonstration vehicle was built based on the BMW Hydrogen 7 bi-fuel car to showcase the emissions reduction potential of a dedicated hydrogen vehicle.

On the hardware side, the most signiﬁcant changes are the removal of the gasoline fuel system including fuel injectors, fuel lines, charcoal ﬁlters for tank ventilation and fuel

(27)

Figure 2.8: Hydrogen Toyota Prius

rail. The two high-pressure fuel pumps were also removed, which reduce the parasitic losses on the engine. For stability reasons, the gasoline fuel tank remains in the vehicle because it is a structural element. The vehicles are equipped with improved catalysts. Independent test results showed that these vehicles achieved emissions levels that were only a fraction of the SULEV standard forN Ox and CO emissions. For non-methane

hydrocarbon (NMHC) emissions, the cycle-averaged emissions were actually 0 g/km, which required the car to actively reduce emissions compared to the ambient concentration. The fuel economy numbers on the FTP-75 test cycle were 3.7 kg of hydrogen per 100 km, which, on an energy basis, is equivalent to a gasoline fuel consumption of 13.8 L per 100 km. Fuel economy numbers for the highway cycle were determined to be 2.1 kg of hydrogen per 100 km, equivalent to 7.8 L of gasoline per 100 km [23].

Ford Motor Company has been evaluating hydrogen since 1997 as an alternative fuel option for vehicles with internal combustion engines. In 2001, Ford presented the hydrogen engine-powered P2000 vehicle, the first production viable, North American OEM hydrogen internal combustion engine vehicle (figure 2.9). The aluminum-intensive five-passenger family sedan was equipped with a highly optimized hydrogen port injection, 14.5:1 compression ratio, 2.0 L engine, gaseousH₂ fuel supply with an operating pressure of up to 250 bar and a triple-redundant hydrogen safety system consisting of gas sensing as well as active and passive elements. The hydrogen P2000 vehicle met SULEV standards for HC and CO and emitted 0.23–0.46 g/km ofN Ox while showing a metro cycle fuel

economy improvement of up to 17.9% relative to gasoline [24].

To demonstrate a commercially viable hydrogen ICE-powered vehicle application, Ford fully engineered a demonstration fleet of 30 E-450 shuttle buses with a 6.8 L Triton engine that runs on hydrogen (figure 2.10). The 8–12 passenger shuttle bus with a 4.5 m wheelbase and an estimated gross vehicle weight of 6373 kg is equipped with a compressed hydrogen on-board storage system that holds up to 29.6 kg of hydrogen at a pressure of 350 bar with a resulting vehicle range of 240–320 km. The target specified for the hydrogen-powered shuttle bus is to meet 2010 Phase II heavy duty emission standards [25]-[28].

(28)

Figure 2.9: Ford P2000

(29)

2.6 Overview of hydrogen vehicles

Table 2.1 summarizes the most relevant information for the hydrogen-powered vehicles that were described in detail in the previous chapters. The summary includes technological aspects such as the type of engine used or the hydrogen storage system as well as vehicle range for hydrogen and, if bi-fuel, gasoline, and the number of vehicles produced or converted.

Name Year Engine Tank Capacity Range Units made

Rivaz 1807 1-cyl Compressed Prototype

Lenoir 1860 1-cyl Water electrolysis Prototype Norsk Hydro 1933 Ammonia reforming Prototype Musashi 1 1974 Compressed 7N m3 _Prototype

Musashi 2 1975 Cryo 230 L Prototype

Musashi 3 1977 2-stroke Cryo 65 L Prototype

BMW 1979 3.5 L Cryo 300 km

Ford P2000 2001 2.0 L I4 Compressed 1.5 kg 100 km BMW Hydrogen 7 2003 6.0 L V12 Cryo 8 kg 200+480 km 100 Mazda RX-8 Hydrogen RE 2003 2x654cc Compressed 2.4 kg 100+550 km >30 Ford Shuttle Bus 2004 6.8 L V10 Compressed 29.6 kg 240÷320 km

ETEC Silverado 2004 6.0 L V8 Compressed 10.5 kg up to 335 km 20 Quantum Prius 2005 1.5 L I4 Compressed 1.6 kg (reg.)/2.4 kg (ext.) 100–130 km >30

(30)

(31)

Chapter 3. Hydrogen properties

Chapter 3 Hydrogen properties

Hydrogen is by far the most plentiful element in the universe, making up 75% of the mass of all visible matter in stars and galaxies. Hydrogen is the simplest of all elements. The hydrogen atom can be visualized as a dense central nucleus with a single orbiting electron, much like a single planet in orbit around the sun. Scientists prefer to describe the electron as occupying a “probability clou” that surrounds the nucleus some-what like a fuzzy, spherical shell.

In most hydrogen atoms, the nucleus consists of a single proton, although a rare form (or “isotope”) of hydrogen contains both a proton and a neutron. This form of hydrogen is called deuterium or heavy hydrogen. Other isotopes of hydrogen also exist, such as tritium with two neutrons and one proton, but these isotopes are unstable and decay radioactively. Most of the mass of a hydrogen atom is concentrated in its nucleus. In fact, the proton is more than 1800 times more massive than the electron. Neutrons have almost the same mass as protons. However, the radius of the electron’s orbit, which deﬁnes the size of the atom, is approximately 100,000 times as large as the radius of the nucleus! Clearly, hydrogen atoms consist largely of empty space. Atoms of all elements consist largely of empty space, although all others are heavier and have more electrons. A proton has a positive electrical charge, and an electron has a negative electrical charge. Neutrons do not carry a charge. Together, the charges associated with the proton and electron of each hydrogen atom cancel each other out, so that individual hydrogen atoms are electrically neutral. Chemically, the atomic arrangement of a single electron orbiting a nucleus is highly reactive. For this reason, hydrogen atoms naturally combine into molecular pairs (H₂in-stead ofH). To further complicate things, each proton in a hydrogen pair has a

ﬁeld associated with it that can be visualized and described mathematically as a “spin”. Molecules in which both protons have the same spin are known as “Orthohydrogen”. Molecules in which the protons have opposite spins are known as “Parahydrogen”. Over 75% of normal hydrogen at room temperature is orthohydrogen. This diﬀerence becomes important at very low temperatures since orthohydrogen becomes unstable and changes to the more stable parahydrogen arrangement, releasing heat in the process. This heat can complicate low temperature hydrogen processes, particularly liquefaction.

(32)

3.1 Comparison with other fuels

It is natural to compare hydrogen to other hydrocarbon fuels with which we are more familiar. All hydrocarbon fuels are molecular combinations of carbon and hydrogen atoms. There are thousands of types of hydrocarbon compounds, each with a speciﬁc combination of carbon and hydrogen atoms in a unique geometry.

The simplest of all hydrocarbons is methane, which is the principal constituent of natural gas. Other components of natural gas include ethane, propane, butane and pentane as well as impurities. Methane has the chemical formula CH₄, which means that each molecule has four hydrogen atoms and one carbon atom.

Other common hydrocarbons are ethane (C₂H₆), propane (C₃H₈) and butane (C₄H₁₀). These are all considered light hydrocarbons since they contain less than ﬁve carbon atoms per molecule and therefore have low molecular weight (a carbon atom is almost 12 times as heavy as a hydrogen atom).

Gasoline and Petrol are composed of a mixture of many diﬀerent hydro-carbons, but an important constituent is heptane (C₇H₁₆). Petrol, diesel, kerosene, and compounds found in asphalt, heavy oils and waxes, are considered heavy hydrocarbons as they contain many carbon atoms per molecule, and therefore have high molecular weight.

The lightest hydrocarbons are gases at normal atmospheric pressure and temperature. Heavier hydrocarbons, with 5 to 18 carbon atoms per compound, are liquid at ambient conditions and have increasing viscosity with molecular weight.

Other chemical fuels include alcohols whose molecules combine an oxygen/hydrogen atom pair (OH) with one or more hydrocarbon groups. Common alcohol fuels are methanol

(CH₃OH) and ethanol (C₂H₅OH). These may be blended with hydrocarbons for use in

internal combustion engines.

Figure 3.1 shows the chemical structure of common fuels.

3.2 Physical and chemical properties of hydrogen

rele-vant to engines

Starting from some physical and chemical properties of hydrogen and hydrogen–air mixtures, a number ofH2ICE features can already be deﬁned or expected.

Table 3.1 lists some properties of hydrogen compared to methane and iso-octane [29]-[32], which are taken here as representing natural gas and gasoline, respectively, as it is easier to deﬁne properties for single-component fuels. The small and light hydrogen molecule is very mobile (high mass diﬀusivity) and leads to a very low density at atmospheric conditions (data given at 300 K and 1 atm).

Table 3.2 shows the vapor and liquid densities of hydrogen, methane and iso-octane. Table 3.3 lists the properties of hydrogen–air mixtures, at stoichiometric and at the lean limit mentioned above, compared to stoichiometric methane–air and iso-octane–air mixtures [29]-[32]. Data given at 300 K and 1 atm (with the exception of the laminar

(33)

17 Chapter 3. Hydrogen properties

Figure 3.1: Chemical structure of common fuels

burning velocity, given at 360 K and 1 atm). The volume fraction of fuel in the fuel–air mixture can be directly calculated from the molar stoichiometric air-to-fuel ratio listed in table 3.1. The large volume fraction occupied by hydrogen has consequences for the attainable engine power density.

Combined with the wide flammability limits, it also has an important effect on mixture properties such as the kinematic viscosity, thermal conductivity, etc. These properties vary much more than in conventionally fueled engines. This affects, for example, non-dimensional numbers used in modeling, such as Reynolds numbers, which can substantially differ from the numbers for hydrocarbon combustion. The comparatively large variation in mixture density and thus, the speed of sound, affects the gas dynamics in engines with external mixture formation.

An increased ratio of specific heat results in an increased amount of compression work. However, the actual compression work, particularly for direct-injection operation, strongly depends on the injection strategy. Calculations have shown that injection timing and duration are the dominating factors compared to the fuel properties, and efficiency benefits of up to 4% can be gained when employing an optimized injection strategy [33].

In engine testing, both the air mass ﬂow rate ˙ma and the fuel mass ﬂow rate ˙mf are

normally measured. The ratio of these ﬂow rates is useful in deﬁning engine operating conditions:

(34)

Property Hydrogen Methane Iso-octane

Molecular weight[g/mol] 2.016 16.043 114.236

Density[kg/m3] 0.08 0.65 692

Mass diﬀusivity in air[cm2/s] 0.61 0.16 0.07

Minimum ignition energy[mJ] 0.02 0.28 0.28

Minimum quenching distance[mm] 0.64 2.03 3.5

Flammability limits in air[vol%] 4-75 5-15 1.1-6

Flammability limits[λ] 10-0.14 2-0.6 1.51-0.26

Flammability limits[φ] 0.1-7.1 0.5-1.67 0.66-3.85

Lower heating value[MJ/kg] 120 50 44.3

Higher heating value[MJ/kg] 142 55.5 47.8

Stoichiometric air-to-fuel ratio[kg/kg] 34.2 17.1 15.0 Stoichiometric air-to-fuel ratio[kmol/kmol] 2.387 9.547 59.666 Table 3.1: Hydrogen properties compared with methane and iso-octane properties

Property Hydrogen Methane Gasoline Vapor density (293 K, 1 atm)[kg/m3] 0.08 0.65 4.4 Liquid density (at normal boiling point, 1 atm)[kg/m3_{] 70.8} _422.8 ₇₀₀

Table 3.2: Vapor and liquid densities of comparative substances

Property H2− air H2− air CH4− air C8H18− air

λ= 1 λ= 4 λ= 1 λ= 1 φ= 1 φ= 0.25 φ = 1 φ= 1 Volume fraction[%] 29.5 9.5 9.5 1.65 Mixture Density[kg/m3] 0.85 1.068 1.123 1.229 Kinematic viscosity[mm2/s] 21.6 17.4 16 15.2 Auto-ignition temperature[K] 858 >858 813 690 Adiabatic flame temperature[K] 2390 1061 2226 2276 Thermal conductivity[10−2W/mK] 4.97 3.17 2.42 2.36 Thermal diffusivity[mm2/s] 42.1 26.8 20.1 18.3 Ratio of specific heats 1.401 1.400 1.354 1.389 Speed of sound[m/s] 408.6 364.3 353.9 334.0 Air-to-fuel ratio[kg/kg] 34.2 136.6 17.1 15.1 Mole ratio before/after combustion 0.86 0.95 1.01 1.07 Laminar burning velocity, 360 K[cm/s] 290 12 48 45 Gravimetric energy content[kJ/kg] 3758 959 3028 3013 Volumetric energy content[kJ/m3] 3189 1024 3041 3704

(35)

AF R= Air/fuel ratio (A/F ) = ˙ma

˙mf

(3.1)

F AR= F uel/air ratio (F/A) = ˙mf

˙ma

(3.2) A stoichiometric ratio is the amount where in the reaction all reagent is consumed, there is no shortfall of reagent, and no residues remain. For hydrogen, the stoichiometric proportions of fuel and air is:

A F S = F A −1 S = 34.2 (3.3)

Because the composition of the combustion products is significantly different for fuel-lean and fuel-rich mixtures, and because the stoichiometric fuel/air ratio depends on fuel composition, the ratio of the actual fuel/air ratio to the stoichiometric ratio is a more informative parameter for defining mixture composition. The fuel/air equivalence ratioφ

is deﬁned as:

φ=(F/A)actual

(F/A)S

(3.4) The inverse ofφ, the relative air/fuel ratio λ is deﬁned as:

λ= φ−1=(A/F )actual

(A/F )S

(3.5) For fuel-lean mixtures: φ <1, λ > 1

For stoichiometric mixtures: φ= λ = 1

For fuel-rich mixtures: φ >1, λ < 1

3.2.1 Flammability Range

The flammability range of a gas is defined in terms of its lower flammability limit (LFL) and its upper flammability limit (UFL). The LFL of a gas is the lowest gas concentration that will support a self-propagating flame when mixed with air and ignited. Below the LFL, there is not enough fuel present to support combustion; the fuel/air mixture is too lean. The UFL of a gas is the highest gas concentration that will support a self-propagating flame when mixed with air and ignited. Above the UFL, there is not enough oxygen present to support combustion; the fuel/air mixture is too rich. Between the two limits is the flammable range in which the gas and air are in the right proportions to burn when ignited.

A stoichiometric mixture occurs when oxygen and hydrogen molecules are present in the exact ratio needed to complete the combustion reaction. If more hydrogen is available than oxygen, the mixture is rich so that some of the fuel will remain un-reacted although all of the oxygen will be consumed. If less hydrogen is available than oxygen, the mixture is

(36)

lean so that all the fuel will be consumed but some oxygen will remain. Practical internal combustion and fuel cell systems typically operate lean since this situation promotes the complete reaction of all available fuel. One consequence of the UFL is that stored hydrogen (whether gaseous or liquid) is not flammable while stored due to the absence of oxygen in the cylinders. The fuel only becomes flammable in the peripheral areas of a leak where the fuel mixes with the air in sufficient proportions.

Two related concepts are the lower explosive limit (LEL) and the upper explosive limit (UEL). These terms are often used interchangeably with LFL and UFL, although they are not the same. The LEL is the lowest gas concentration that will support an explosion when mixed with air, contained and ignited. Similarly, the UEL is the highest gas concentration that will support an explosion when mixed with air, contained and ignited.

An explosion is different from a fire in that for an explosion, the combustion must be contained, allowing the pressure and temperature to rise to levels sufficient to violently destroy the containment. For this reason, it is far more dangerous to release hydrogen into an enclosed area (such as a building) than to release it directly outdoors.

The wide range of ﬂammability limits for the hydrogen, with ﬂammable mixtures from as lean asλ= 10 to as rich as λ = 0.14 (0.1 < φ < 7.1) allows a wide range of engine power

output through changes in the mixture equivalence ratio. The flammability limits widen with increasing temperature, with the lower flammability limit dropping to 2 vol% at 573 K (equivalent toλ= 20, φ = 0.05) (figure 3.2). The lower flammability limit increases

with pressure [34], with the upper ﬂammability limit having a fairly complex behavior in terms of pressure dependence [35] but of lesser importance to engines.

Figure 3.3 shows the ﬂammability ranges of comparative fuels at atmospheric tempera-ture.

(37)

Figure 3.3: Flammability ranges of comparative fuels at atmospheric temperature

The lean limit of H2ICEs is reached for lower air-to-fuel equivalence ratios in the

vicinity ofλ= 4, φ = 0.25. The lower ﬂammability limit is mostly determined by the

classical method of flame propagation in a tube. The mass diffusivity of hydrogen is high, and this causes a difference in the limit for upward or downward propagating flames, due to preferential diffusion in the presence of buoyancy [36], [37]. For upward propagating flames, mixtures as lean as 4% hydrogen in air are still flammable but are non-coherent and burn incompletely. The value of 4% pertains to one particular experimental configuration, so in real-world situations, the limit may well be below 4% (or above, depending on conditions) [38]. The absolute limit is thus not well known even today. However, this limit is important for safety considerations but less so for engine combustion.

(38)

3.2.2 Ignition energy

Ignition energy is the amount of external energy that must be applied in order to ignite a combustible fuel mixture. Energy from an external source must be higher than the auto-ignition temperature and be of suﬃcient duration to heat the fuel vapor to its ignition temperature. Common ignition sources are ﬂames and sparks.

The minimum ignition energy of a hydrogen–air mixture at atmospheric conditions is an order of magnitude lower than for methane–air and iso-octane–air mixtures. It is only 0.017 mJ, which is obtained for hydrogen concentrations of 22–26% (λ= 1.2 − 1.5, φ= 0.83−0.67) [39]. The minimum ignition energy is normally measured using a capacitive

spark discharge, and thus is dependent on the spark gap. Using a 2 mm gap, the minimum ignition energy is about 0.05 mJ and more or less constant for hydrogen concentrations between 10% and 50% (λ= 0.42 − 3.77, φ = 2.38 − 0.27), with a sudden increase when

the concentration ofH2 is below 10% [39].

3.2.3 Quenching distance

The quenching gap (or quenching distance) describes the flame extinguishing properties of a fuel when used in an internal combustion engine. Specifically, the quenching gap relates to the distance from the cylinder wall that the flame extinguishes due to heat losses. The quenching distance can be experimentally derived from the relation between the minimum ignition energy and the spark gap size [40] or directly measured [41]. It is minimal for mixtures around stoichiometry, and decreases with increasing pressure and temperature. The quenching gap of hydrogen (0.64 mm) is approximately 3 times less than that of other fuels, such as wall before they are extinguished making them more difficult to quench than gasoline flames. As can be seen in Table 3.1, it is about one-third that for methane and iso-octane.

This smaller quenching distance can also increase the tendency for backfire since the flame from a hydrogen-air mixture can more readily get past a nearly closed intake valve than the flame from a hydrocarbon-air mixture. This affects crevice combustion and wall heat transfer.

3.2.4 Energy content

Every fuel can liberate a fixed amount of energy when it reacts completely with oxygen to form water. This energy content is measured experimentally and is quantified by a fuel’s higher heating value (HHV) and lower heating value (LHV). The difference between the HHV and the LHV is the “heat of vaporization” and represents the amount of energy required to vaporize a liquid fuel into a gaseous fuel, as well as the energy used to convert water to steam.

The higher and lower heating values of comparative fuels are indicated in table 3.4. Gaseous fuels are already vaporized so no energy is required to convert them to a gas.

(39)

Fuel Higher heating value Lower heating value (at 1 atm and 298 K) (at 1 atm and 298 K)

[kJ/g] [kJ/g] Hydrogen 141.86 119.93 Methane 55.53 50.02 Propane 50.36 45.6 Gasoline 47.5 44.5 Diesel 44.8 42.5 Methanol 19.96 18.05

Table 3.4: Higher and lower heating values of comparative fuels

The water that results from both a combustive reaction and the electrochemical reaction within a fuel cell occurs as steam, therefore the lower heating value represents the amount of energy available to do external work.

Both the higher and lower heating values denote the amount of energy (Joules) for a given weight of fuel (kilograms). Hydrogen has the highest energy-to-weight ratio of any fuel since hydrogen is the lightest element and has no heavy carbon atoms. It is for this reason that hydrogen has been used extensively in the space program where weight is crucial. Speciﬁcally, the amount of energy liberated during the reaction of hydrogen, on a mass basis, is about 2.5 times the heat of combustion of common hydrocarbon fuels (gasoline, diesel, methane, propane, etc.) Therefore, for a given load duty, the mass of

hydrogen required is only about a third of the mass of hydrocarbon fuel needed.

Whereas the energy content denotes the amount of energy for a given weight of fuel, the energy density denotes the amount of energy (Joules) for a given volume (inm3) of fuel. Thus, energy density is the product of the energy content (LHV in our case) and the density of a given fuel.

The energy density is really a measure of how compactly hydrogen atoms are packed in a fuel. It follows that hydro-carbons of increasing complexity (with more and more hydrogen atoms per molecule) have increasing energy density. At the same time, hydrocarbons of increasing complexity have more and more carbon atoms in each molecule so that these fuels are heavier and heavier in absolute terms.

On this basis, hydrogen’s energy density is poor (since it has such low density) although its energy to weight ratio is the best of all fuels (because it is so light). The energy density of comparative fuels, based on the LHV, is indicated in table 3.5.

3.2.5 Flashpoint

All fuels burn only in a gaseous or vapor state. Fuels like hydrogen and methane are already gases at atmospheric conditions, whereas other fuels like gasoline or diesel that are liquids must convert to a vapor before they will burn. The characteristic that describes how easily these fuels can be converted to a vapor is the flashpoint. The flashpoint is defined as the temperature at which the fuel produces enough vapors to form an ignitable mixture with air at its surface.

(40)

Fuel Energy density (LHV)

Hydrogen 10,050kJ/m3; gas at 1 atm and 288 K 1,825,000kJ/m3; gas at 200 barg and 288 K 4,500,000kJ/m3; gas at 690 barg and 288 K 8,491,000kJ/m3; liquid

Methane 32,560kJ/m3; gas at 1 atm and 288 K 6,860,300kJ/m3; gas at 200 barg and 288 K 20,920,400kJ/m3; liquid

Propane 86,670kJ/m3; gas at 1 atm and 288 K 23,488,800kJ/m3; liquid

Gasoline 31,150,000kJ/m3; liquid

Diesel 31,435,800kJ/m3 minimum; liquid Methanol 15,800,100kJ/m3; liquid

Table 3.5: Energy densities of comparative fuels

If the temperature of the fuel is below its flashpoint, it can-not produce enough vapors to burn since its evaporation rate is too slow. Whenever a fuel is at or above its flashpoint, vapors are present. The flashpoint is not the temperature at which the fuel bursts into flames; that is the auto-ignition temperature.

The flashpoint is always lower than the boiling point. For fuels that are gases at atmospheric conditions (like hydrogen, methane and propane), the flashpoint is far below ambient temperature and has little relevance since the fuel is already fully vaporized. For fuels that are liquids at atmospheric conditions (such as gasoline or methanol), the flash-point acts as a lower flammability temperature limit.

Table 3.6 shows the ﬂashpoint of comparative fuels.

Fuel Flashpoint Hydrogen <20 K Methane 85 K Propane 169 K Gasoline Approximately 230 K Methanol 284 K

Table 3.6: Flashpoint of comparative fuels

3.2.6 Auto-ignition temperature

The auto-ignition temperature is the minimum temperature required to initiate self-sustained combustion in a combustible fuel mixture in the absence of a source of ignition. In other words, the fuel is heated until it bursts into ﬂame.

There is some ambiguity concerning the autoignition temperature of fuels in general and hydrogen in particular (table 3.7). For instance, for methane values have been found ranging from 810 K [42] to 868 K [43]. For hydrogen, values were found from 773 K [44] to 858 K [45]. Some sources list the autoignition temperature for hydrogen as lower than

(41)

25 Chapter 3. Hydrogen properties that for methane; other sources list the opposite. This ambiguity can be at least partly explained by the sensitivity of autoignition temperatures to the experimental apparatus, the experimental procedure and the criterion used for deﬁning the value [46].

Fuel Auto-ignition temperature Hydrogen 773 - 858 K

Methane 810 - 868 K Propane about 750 K Methanol about 660 K Gasoline 500 - 750 K

Table 3.7: Auto-ignition temperature of comparative fuels

For spark-ignition engines, with a propagating flame front, autoignition of the unburned mixture ahead of the flame front is unwanted, as it can result in knocking combustion. The efficiency of a spark-ignition engine is influenced by the compression ratio and the ignition timing (among others), the choices of which are dependent on the autoignition temperature of the fuel–air mixture, so this is an important parameter.

For liquid hydrocarbons, the octane rating is more commonly used as a measure of the propensity of a fuel–air mixture to undergo pre-ﬂame reactions. The octane number describes the anti-knock properties of a fuel when used in an internal combustion engine. Knock is a secondary detonation that occurs after fuel ignition due to heat buildup in some other part of the combustion chamber. When the local temperature exceeds the auto-ignition temperature, knock occurs. The performance of the hydrocarbon octane is used as a standard to measure resistance to knock, and is assigned a relative octane rating of 100. Fuels with an octane number over 100 have more resistance to auto-ignition than octane itself.

Hydrogen has a very high research octane number and is therefore resistant to knock even when combusted under very lean conditions. The octane number of comparative fuels are indicated in Table 3.8. For hydrogen, a research octane number (RON) in excess of 130 and a motor octane number (MON) of 60 are shown [47]-[49].

Fuel Octane Number

Hydrogen 130+ (lean burn)

Methane 125

Propane 105

Octane 100

Gasoline 87

Diesel 30

Table 3.8: Octane number of comparative fuels

3.2.7 Laminar burning velocity

The laminar burning velocity of a fuel–air mixture is an important physicochemical property due to its dependence on pressure, temperature, mixture equivalence ratio and diluent

(42)

concentration. It affects the combustion rate in an engine, the equivalence ratio limits for stable combustion, the tolerance for EGR, etc. Most engine combustion models assume the flame structure to be that of a (stretched) laminar flame, with the effect of the in-cylinder turbulence to be one of stretching and wrinkling the flame, thereby increasing the flame area. Consequently, data on the laminar burning velocity and its dependence on pressure, temperature, mixture composition and stretch rate are a prerequisite [50].

The laminar burning velocity of stoichiometric hydrogen–air mixtures is much higher than that of methane and iso-octane. However, if lean-burn strategies are used, the burning velocity can be much lower (see value forλ= 4, φ = 0.25). For mixtures around

stoichiometry, the high burning velocity and high adiabatic ﬂame temperature point to high nitrogen oxides (N Ox) emissions.

(43)

Chapter 4. Problem deﬁnition

Chapter 4 Problem deﬁnition

The use of hydrogen is foreseen as one of the solutions useful to decrease air urban pollution, owing to the potential zero or near zero local emissions when used as a fuel for vehicles. While we are looking at fuel cells at least as a long-term solution for hydrogen eﬃcient use, hydrogen fueled internal combustion engines still constitute a viable short or mid-term alternative.

It is widely recognized that an internal combustion engine loses nearly half of its power when carburetted with hydrogen, as well as its tendency to backfire and pre-ignition [51], [52]. This large loss of delivered power is due to several factors: the first is the loss in volumetric efficiency owing to the low density or hydrogen; other factors are the tendency to backfire and pre-ignite when approaching stoichiometric operations that in fact limit the use of much lower values of the equivalence ratio than unity [53].

Therefore, an eﬃcient use of hydrogen in internal combustion engines requires the use of direct-injection techniques. This is especially true when injecting after Inlet Valve Closing (IVC); in this case, backﬁre cannot happen and moreover hydrogen enters the cylinder after the intake phase that in turn operates quite a strong cooling of the metallic walls.

Hydrogen direct injection implicates the use of high injection pressures, owing to the low density of hydrogen; in turn, this leads to the use of high-pressure tanks to reach a suﬃcient operating vehicle range.

The ﬁnal goal of this research is the development of a prototype engine fueled with directly injected hydrogen at low pressure, about 10-20 bar. In particular, in this phase a speciﬁc engine and injector type were chosen.

4.1 The Lombardini 15LD500 engine

The main baseline engine requirements can be summarized as a suﬃcient space availability to locate one or two hydrogen injectors, one piezoelectric pressure sensor and the spark plug, an appropriate turbulence motion in the cylinder to favor the mix between air and hydrogen and a reliable and robust structure to support high pressure peaks due to

(44)

abnormal combustion that may occur during the experiments.

For this study, the Lombardini 15LD500 was used; it’s a four-stroke, single cylinder, naturally aspirated, air cooled diesel engine (figure 4.1). Its main features are: direct injection with mechanical pump and five holes injector, forced lubrication with oil pump, internal full flow oil filter, oil breathing blow-by with safety device, die-cast aluminum crankcase, re-borable independent cast-iron cylinders and aluminum alloy cylinder head.

Figure 4.1: Lombardini 15LD500 engine

A compression ignition engine was used for this experiment instead of a spark ignition one taking into consideration the main features of the hydrogen combustion itself. Hydrogen often causes near-constant volume combustion, very similar to the ﬁrst phase of diesel combustion process. Moreover, hydrogen allows throttle operation at partial load too, in a very similar way to compression ignition engines. This engine also has a very simple head owing to the presence of only two valves and to the lack of the water jackets. Finally, the highly swirled intake system was considered favorable for mixing between air and fuel.

The speciﬁcations of the Lombardini 15LD500 are shown in table 4.1.

Figure 4.2 shows the power curves, the torque curves and the speciﬁc fuel consumption curve. Figure 4.3 shows the technical drawings of the engine.

4.2 The hydrogen injection system

The ﬁrst analysis step was the evaluation of the engine air capacity in order to calculate the eﬀective need of fuel mass per cycle. On this basis, the choice for the most suitable

(45)

29 Chapter 4. Problem deﬁnition Cylinders N. 1 Displacement cm3 505 Bore mm 87 Stroke mm 85 Compression Ratio 19:1

RatingkW/HP N (80/1269/CEE) ISO 1585 8.8/12.0 NB ISO 3046 IFN 8.2/11.1 NA ISO 3046 ICXN 7.5/10.2

Max. torque N m 30.0@2200rpm

Min. idling speed rpm 1200

Fuel tank capacity L 5

Oil consumption kg/h 0.0055

Oil sump capacity L 1.5

Min allowable oil pressure bar 0.8 Max allowable inclination for:

- short periods of operation (max time 30 minutes) deg 25 - peak values (max time 1 minute) deg 35 Cap. of air required for correct combustion @3600rpm L/min 800 Cap. of air required for correct cooling @3600rpm L/min 8700

Dry weight kg 48

Table 4.1: Lombardini 15LD500 speciﬁcations

injector was performed, trying to keep hydrogen injection pressure as low as possible. The latter is a fundamental issue; in fact in the case of the use of pressurized tanks, the lowest is the injection pressure, the longest is the vehicle operating range between two refuellings. Furthermore, a low injection pressure makes use of metal hydride tanks viable for vehicle operations [54].

The intake system, composed of the air filter, intake manifold, intake port and intake valve restricts the amount of air which an engine of a given displacement can induct. The parameters used to measure the effectiveness of an engine’s induction process is the volumetric efficiencyηv.

Volumetric efficiency is only used with four-stroke cycle engines which have a distinct induction process. It is defined as the volume flow rate of air into the intake system divided by the rate at which volume is displaced by the piston

ηv= 2 ˙m a

ρa,iVdN

(4.1) where ˙ma is the air mass ﬂow rate,ρa,iis the inlet air density (1.18kg/m3 at 25°C and

1 atm),Vd is the displacement andN is the rpm.

Volumetric efficiency is affected by fuel type, fuel/air ratio, fraction of fuel vaporized in the intake system, fuel heat of vaporization, mixture temperature as influenced by heat transfer, ratio of exhaust to inlet manifold pressure, compression ratio, engine speed, intake and exhaust manifold and port design, intake and exhaust valve geometry, size, lift and timings.

Typical maximum values ofηv for naturally aspirated engines are in the range 70 to 90

percent. The volumetric eﬃciency for diesel is somewhat higher than for SI engines. In this ﬁrst analysis, aηv= 0.8 was chosen in order to evaluate the engine air capacity.

(46)

Figure 4.2: Lombardini 15LD500 engine - Curves

The maximum speedN = 3000rpm (50giri/s) was chosen for this ﬁrst calculation. The

following models take into account more speciﬁc conditions in terms of air and hydrogen mass ﬂow rate.

The air mass ﬂow rate in the worst condition ˙ma was:

˙ma= ηvVdN ρa,i 2 = 0.8 · 0.000505 · 50 · 1.18 2 = 0.012 kg s (4.2)

For each speed of the engine, the aspirated air in cylinder for each cycle was calculated throughma= ˙mat720 wheret720= 2 · 60/N is the time to have a cycle (720 degrees).

ma= ˙mat720 = ˙ma2 · 60/N = 0.012 · 2 · 60/3000 = 0.00048

kg

cycle= 0.48 g

(47)

31 Chapter 4. Problem deﬁnition

Figure 4.3: Lombardini 15LD500 engine - Technical drawings

In order to study the hydrogen injection varying with injection pressure and rpm of the engine, an equivalence ratioφ was ﬁxed to 0.8 ((AF R)actual= 42.75).

The hydrogen engine is operated without throttling; so the equivalence ratio corresponds to the engine load. Some experimental tests show that the maximum brake thermal eﬃciency was obtained with equivalence ratio near 0.8. Moreover, a full load working condition was considered taking into consideration the worst operating situation in order to choose the injection system [56].

The hydrogen necessary for each cycle was valuated as

mH2= ma (AF R)actual = 0.48 42.75 = 0.011 g cycle (4.4)

In ﬁgure 4.4, the air and hydrogen mass ﬂow rates obtained in the preliminary study were shown.

The time available for injection of gaseous fuels is significantly shorter than that available for injection of liquid fuels. In order to avoid inlet air manifold back fire, the hydrogen injection may be limited during the compression stroke. Constraining the injection of gaseous fuel to the compression stroke avoids displacement of air from the engine cylinder but significantly reduces the time available for injection, increasing the flow rate required to meter the fuel required for full load operation [51], [57].

The choice of very low hydrogen injection pressure requires the adaptation of an injector with a very large cross-sectional area. Thus this study is based on Strata I commercial injector, realized by Synerject Company. The lowest working injection pressure chosen

(48)

Figure 4.4: Preliminary study - Air and hydrogen mass ﬂow rate was 13 barA [58], [59].

This injector was initially developed for air-assisted gasoline direct injection and in this application the nozzle used for the air/fuel rich mixture injection was used to introduce the hydrogen into the engine.

However, one injector per cylinder does not provide a suﬃcient cross-sectional area and therefore two injectors per cylinder were considered.

Since the injection pressure chosen is greater than the regular operating pressure of Synerject injectors, a Bigas valve connected upstream of the Strata injector was taking into consideration to prevent hydrogen leakages.

Thanks to the large ﬂow area of the air assisted Synerject injectors, this kind of injector was considered to be suitable for this application; moreover, it allows the regulation of the necessary mass of hydrogen, for each load and for each speed.

(49)

Chapter 5. Numerical models

Chapter 5 Numerical models

The transformation of an engine for hydrogen fueling is a problem that involves diﬀerent aspects: on one hand, we must take into account the hydrogen mass needed per cycle as a function of rpm, load, injection pressure and, ﬁnally, injection timing. On the other hand, it was considered to be useful the knowledge of hydrogen distribution within the cylinder when ignition occurs.

In order to investigate all these aspects, two one-dimensional engine models were created, both for diesel and hydrogen engines. The one-dimensional models were used to model the global behavior of the engine in terms of thermal balance, mechanical loss, indicated pressure and air capacity and to analyze the hydrogen injection phase.

Moreover, in order to evaluate the capability of the injection system, a zero-dimensional model of the entire injection group was created. With this model, the dynamic and ﬂuid-dynamic behavior of the injectors, as function of timing and injection pressure were analyzed.

5.1 Diesel engine model

A one-dimensional diesel engine model was created by means of the LMS AMESim code (figure 5.1) using mechanical, logical and engine components libraries. This model was built and calibrated to correctly represent the engine in terms of heat exchange, indicated cycle, thermal and mechanical loss. The variations in mechanical losses and heat exchange phenomena arising from the conversion from diesel to hydrogen operations were not considered to be significantly influential on the behavior of the engine as regards the equivalent politropic index of the compression phase. In fact, as this engine is a direct injection one, hydrogen injection takes places mainly during the compression stroke. Mechanical friction, discharge coefficients in the intake and exhaust system and thermal exchange coefficients were considered to be equal for diesel and hydrogen engine.

The cylinder and the combustion chamber were modelled in a zero-dimensional way, while charge motions inside intake and exhaust ducts were modelled in a one-dimensional way.

(50)

Figure 5.1: 1D diesel engine - One-dimensional model

In this model, piston and valves motion were taken into account, as well as heat exchange phenomena, fuel injection and combustion. The heat exchange was modelled on the Woschni model by deﬁning the wall temperature of piston, liner and cylinder head. The mechanical loss was modelled by means of a friction mean eﬀective pressure sub model [55].

The use of these models required a validation that was performed through the compar-ison between the results of the experimental and numerical data.

An experimental session was performed by testing the engine chosen at the engine test bench of the Laboratory of the Department of Energetica. Torque, power, fuel consumption, indicated cycle and heat release rate were measured.

The engine was equipped with a piezoelectric pressure sensor Kistler 6052, a shaft encoder with 720 steps per revolution and several thermocouples to measure intake and exhaust temperatures, as well as to check the thermal stability of the engine during the test. A shielded thermocouple was used to measure the exhaust temperature.

The indicated cycle was measured by averaging 50 cycles for every test condition. This value was chosen in consequence to a sensitivity study, which stated that 50 cycles are suﬃcient to avoid the inﬂuence of cyclic dispersion.

The subsequent figures show the comparison between experimental data and numerical results as regards the brake torque (figure 5.2), the brake power (figure 5.3), the exhaust gas temperature (figure 5.4) and the brake specific fuel consumption BSFC (figure 5.5).

Numerical and experimental results adherence was considered to be satisfactory to represent the engine global behavior during regular diesel operation. As stated before,